Thinking is occurring

I was thinking while shopping on Friday evening, to the extent that I walked away from the cashier without paying - who came out after me as I wandered absent-mindedly across the parking lot sipping on a chocolate steri-stumpi (with the unfortunate effect of breaking my train of thought). My thoughts were continuing from when Jerith mentioned some weeks ago a problem with Descarte's "I think therefore I am": that it should really be "Thinking is occurring, therefore something is". If reality is furthermore what it appears to be (not a simulation/deception), then the thinking is occuring within the head of the creature through which which sensory input is entering those thoughts. The creature in which these thoughts occur prior to being written down, is here being indicated by the first-person personal pronoun, "I". This gets around what "I" means, but leaves open the question of what it means to think. Descarte's statement relates to self-awareness: the phenomenon of thoughts acknowledging that thinking is occurring. Douglas Hofstatder, in "Godel, Escher, Bach: An Eternal Golden Braid", considers self (or self-awareness) to be a "strange loop", a vaguely-defined phenomenon where something recurses infinitely by self-reference yet ends up where it was, along the lines of the Penrose stairs, the ever-increasing scale of notes, or Godel's first incompleteness theorem (where a number-theoretical statement G, encoded as a number, asserts G's own un-provability by means of number theory). The word "self", though, is nothing special, simply referring (like "I") to this instance of a human being having the thought as opposed to some other instance. It has been said that when we have precise definition of what it means to think, we will also be able to develop a computer that thinks - and is perhaps aware of it. As for whether computers think already, Dijkstra compared it to asking whether submarines swim: the computation of a machine learning algorithm and the deliberation in human biology are qualitatively different. Computation in humans is almost entirely below the level of thought, occurring in neurons and brain regions that the resulting thoughts are entirely unaware of. Thoughts can also do computation, if inefficiently - just consider the vast amount of computation behind even a simple thought, and then the amount of conscious thought required to simulate even a simple artificial neuron. Human thought also appears tightly linked with the human capacity for language (another favourite topic of Hofstatder's). To me, language is the most striking distinction between humans and other higher animals, the facility that grants us an almost unlimited ability to manipulate and communicate abstractions and symbols (although imperfectly: there are many cognitive biases). Other animals are probably mostly limited to the space of images and concrete objects - what would remain if you managed to purge words and symbols from your thoughts? Those were the handful of thoughts in my head last Friday evening. (Note: I searched on terms to link to Wikipedia after the fact - should I rather be leave them out? That way people could search on whatever they want, and it doesn't look like I got everything off the wiki.)

Umeshisms: better strategies for life

If you’ve never missed a flight, you’re spending too much time in airports. Scott Aaronson takes the principle of Umeshisms to be "concentrate on the high order bits" - don't have your life dominated by the effort of avoiding minor negative events, because the optimal strategy doesn't avoid those events completely. "If you have never regretted a blog entry, your blog is boring". I was thinking about them occasionally the last few weeks, and have tried to come up with decent general form. It goes like this: If you never experience negative event X, you would be better off with a different approach, even though that approach means X might happen sometimes. From that template you can generate all sorts of wisdom from the ages:

• If you never get rejected, you aren't asking enough (Scott Aaronson).
• A person who never made a mistake never tried anything new (Einstein).
• If you never fail, you aren't aiming high enough.

The following I started applying to my own life a couple of months ago: if you've never forgotten to lock your room, you're spending too much time going back to check.

"What is mind?" and the software analogy

Mind is often compared to the software running on the hardware of the brain. It is a process - as has been said, "Asking where the mind goes when one dies is like asking where the 70-miles-per-hour went in an auto crash". Many however hold dualistic beliefs - that mind is separate from the brain and can exist after brain-death. Dualism is normally part of a set of supernatural beliefs, but is probably argued from a lack of understanding of how the mind results from activities in the brain (from ignorance, that is). Which is true enough: nobody does understand precisely how a network of neurones results in qualitative perceptions and an awareness of "self". What we do have on hand are observations of neural correlates of emotion, perception, etc. Feeling anger, seeing an object, making some ("free will") decision, and behaviors are all correlated with patterns of neural activity. Psychoactive drugs change these patterns in predictable ways while simultaneously changing your perception of reality, and even your personality. Knock someone just hard enough over the head, and the activity stops temporarily (along with the mind). Although neural correlates do not reveal the mind-process, we can infer such a process. Using the computer analogy, imagine observing the chips in a computer with an instrument to detect electric fields. We provide an input that starts a graphics intensive application, and observe higher activity patterns within certain chips on a certain board (that we then identify as some sort of "graphics card"). We fail to observe the algorithm, the "how" behind the vastly complex, detailed, moving images on the display, but we do infer that the associated electrical patterns conceal a process producing the images (does that inference need more justification?). Likewise, the changing neural states correlated with a qualitative experience conceal the process that is the experience. The mind-brain situation today is then like having a computer, a coarse interface to it (through sensory input) and an instrument to measure electric fields in the chips. We can infer that there is a process associated with the changing electrical patterns, but lack the keyboard and step-through debugger needed to observe the process itself.